In previous work, a computational model was derived for describing basilar-membrane displacement in the human ear. The model specifies membrane displacement at arbitrary points for known sound pressure at the eardrum. Its mathematical tractability makes it an attractive tool for digital analysis of speech and other auditory signals. It has, therefore, been approximated in terms of sampled-data functions and programmed for a digital computer. The present simulation represents the motion at 40 points along the membrane. The equivalent spacing of the points is 0.5 mm, and they span the frequency range 75 to 4600 cps. Two different processings of the membrane-displacement information so obtained are described. One is a rectification, smoothing, and power-law amplification of the displacement function at each membrane point. At a given instant of time, therefore. the successive outputs represent values of a short-time amplitude spectrum as produced by the frequency weighting of the membrane. A second processing produces a measure approximating the mean rate of neural firings at successive membrane points. Both processes reflect speech features, such as formant structure, voicing, and friction. and suggest several possibilities for speech analysis.